: NLP Vocabulary: natural language generation

applied NLP > human-computer interaction > task-oriented dialogue > natural language generation

natural language generation

The subfield of artificial intelligence and computational linguistics that is concerned with the construction of computer systems than can produce understandable texts in English or other human languages from some underlying non-linguistic representation of information. (Reiter and Dale, Building applied natural language generation systems, in Natural Language Engineering)

Automatic text generation is a long-standing challenging task as it needs to solve at least three major problems: (1) content selection ("what to say") identifying pertinent information to present (2) text planning ("when to say what") arranging content into ordered sentences and (3) surface realization ("how to say it") deciding words and syntactic structures that deliver a coherent output based on given discourse goals (McKeown 1985). (Hua & Wang, 2019)
Both integrate dialogue context information into NLG for generating more variant and contextual response in task-oriented dialogue systems. (Liu, Maier, Minker & Ultes, 2021)
Natural Language Generation (NLG) for non-English languages is hampered by the scarcity of datasets in these languages. (Kumar, Shrotriya, Sahu, Mishra, Dabre, Puduppully, Kunchukuttan, Khapra & Kumar, 2022)
The emergence of comparative evaluation in NLG raises the broader question of how systems that generate language should be compared. (Belz & Gatt, 2008)
To improve the quality of natural language generation it is crucial to refine the training datasets. (Tsukagoshi, Hirao, Morishita, Chousa, Sasano & Takeda, 2024)

http://data.loterre.fr/ark:/67375/8LP-V83NBD43-5

RDF/XML TURTLE JSON-LD Last modified 10/12/24

Loterre